CVE-2026-22813
The markdown renderer used for LLM responses will insert arbitrary HTML into the DOM.
Does this matter?
High impact if exploited, but EPSS currently rates exploitation as unlikely (0.92%). Schedule it in the normal patch cycle and watch for a rise in EPSS or a public exploit.
Description
OpenCode is an open source AI coding agent. The markdown renderer used for LLM responses will insert arbitrary HTML into the DOM. There is no sanitization with DOMPurify or even a CSP on the web interface to prevent JavaScript execution via HTML injection. This means controlling the LLM response for a chat session gets JavaScript execution on the http://localhost:4096 origin. This vulnerability is fixed in 1.1.10.
- CVSS 4.0
- 9.4 CRITICALCVSS:4.0/AV:N/AC:L/AT:N/PR:N/UI:P/VC:H/VI:H/VA:H/SC:H/SI:H/SA:H/E:X/CR:X/IR:X/AR:X/MAV:X/MAC:X/MAT:X/MPR:X/MUI:X/MVC:X/M
- EPSS
- 0.92% probability · 58th percentile
- CISA KEV
- Not listed
- Weakness
- CWE-79
- Affected
- anoma/opencode
- Source
- security-advisories@github.com
References
- https://github.com/anomalyco/opencode/security/advisories/GHSA-c83v-7274-4vgpExploit, Vendor Advisory
- https://github.com/anomalyco/opencode/security/advisories/GHSA-c83v-7274-4vgpExploit, Vendor Advisory
Source: NVD record, EPSS from FIRST.org, KEV from CISA. Refreshed daily.